Community Detection in Complicated Network Based on the Multi-view Weighted Signed Permanence

Juan-Hui Li, Pei-Zhen Li, Chang-Dong Wang, Jian-Huang Lai

Abstract: This paper addresses the community detection problem in multi-view weighted signed network. Although some methods have been developed to detect communities in multi-view network, they are mostly designed for the unweighted unsigned case. There is still a lack of methods for multi-view weighted signed network. Since an increasing number of multi-view weighted signed networks are being generated in some applications, there is a necessity to develop a community detection approach to reveal the complicated structure of such networks. In this paper, we extend the single-view permanence model to the multi-view weighted signed case, which is a node-level model by considering four factors based on the influence of the nodes, including internal connections, total connections, maximum external connections and internal clustering coefficient. Extensive experiments are conducted on some networks to evaluate the performance of our proposed approach.